Hierarchical communities in the larval Drosophila connectome: Links to cellular annotations and network topology

One of the longstanding aims of network neuroscience is to link a connectome's topological properties-i.e. features defined from connectivity alone--with an organism's neurobiology. One approach for doing so is to compare whole-brain maps of connectome properties with maps of metabolic, functional, and neurochemical annotations. This type of analysis is popular at the meso-/macro-scale, but is less common the nano-scale, owing to a paucity of neuron-level connectome data. However, recent methodological advances have made possible reconstruction of whole-brain connectomes at single-neuron resolution for a select set of organisms. These include the fruit fly, Drosophila melanogaster , and its developing larvae. In addition to fine-scale descriptions of neuron-to-neuron connectivity, these datasets are accompanied by rich annotations, documenting cell type and function. Here, we use a hierarchical and weighted variant of the stochastic blockmodel to detect multi-level communities in a recently published connectome of the larval Drosophila . We find that these communities, which are detected using a data-driven method based on connectivity alone, neatly partition neurons based on function, cell types, and macro-level cell class. We find that communities mostly interact assortatively, reflecting the principle of functional segregation. However, a small number of communities interact non-assortatively. These neurons, which also form a "rich-club", overlap with interneurons that receive and process sensory signals, delivering the outputs to effectors. Finally, we investigate the role of community structure in shaping neuron-to-neuron communication patterns. We explore three hypothetical communication policies, in each case discovering that neurons frequently assigned to the same community across hierarchical levels have greater capacity for communication (reduced path length and increased communicability). Finally, we investigate in granular detail the structure of shortest paths and how they move between communities and hierarchical levels. We find that interneurons and high-degree hubs deviate from a characteristic "u-shaped" trajectory, moving quickly from their own module towards their target via strong inter-modular projections. Our findings recapitulate and extend other recent studies investigating the community structure of the larval Drosophila connectome, while establishing links between communities and neuronal annotations. ### Competing Interest Statement The authors have declared no competing interest.